Summary

This project assessed swathing and unmanned aerial vehicle (UAV) imagery as tools to manage barley grass in low rainfall farming systems. UAV mapping identified weed patches more effectively in pastures than cereals, with higher resolution flights improving accuracy. However, the cost and expertise required may limit adoption.

Early swathing of cereal crops at 20 to 40 per cent of grain moisture captured more weed seed, but 40 to 60 per cent of barley grass seed had already shed by this time. By crop maturity, only 10 to 20 per cent remained. Early harvest is essential to maximise seed collection. Together, UAV imagery and early swathing offer integrated options for managing herbicide-resistant barley grass.

Background

Barley grass is an increasingly problematic weed in low rainfall mixed farming systems, particularly due to early seed drop and emerging herbicide resistance, including confirmed Group 1 resistance. Unlike ryegrass, which retains seed longer, barley grass sheds a large portion before standard harvest timing, limiting the usefulness of harvest weed seed control (HWSC) tools like chaff carts, windrows and harvest weed seed destructors. This project explored the value of UAV imagery to identify infestations in paddocks and assessed whether swathing crops earlier than harvest could improve seed capture and reduce the weed seed bank.

Research Aims

The core objectives of the project were to:

• Investigate barley grass seed retention timing in cereal crops
• Assess the effectiveness of early swathing and HWSC in capturing barley grass seed
• Evaluate UAV imagery as a tool for detecting grass weed infestations
• Explore opportunities for precision weed management based on UAV-derived maps.

In The Field

Research was conducted over three seasons across several paddocks on the Upper Eyre Peninsula. Barley grass presence and seed retention were monitored weekly from early October to harvest. Swathing was trialled when grain moisture was between 20 and 40 per cent to assess barley grass weed seed capture potential.

Simultaneously, UAV (drone) imagery was captured at both 40m and 120m altitudes using visible light and near infra-red sensors. Imagery was cross-referenced with on-ground paddock assessments marked with GPS and visual signs. Software was used to process the imagery to identify high-density weed patches. Analysis also included comparing cereal crops (wheat and barley) and legumes for weed visibility and competition.

Results

At the first swathing opportunity (when grain moisture was 25-40 per cent), 40 to 60 per cent of barley grass seed had already shed. By full maturity, only 10 to 20 per cent remained on the plant, confirming the low seed retention.

Early swathing modestly increased the seed capture potential but did not fully prevent seedbank replenishment.

Drone imagery reliably identified dense weed patches in legumes and pastures but was less effective in cereals unless recorded at high resolution. Imagery at 40m provided better results than 120m, although file size and analysis complexity increased (see images at right). UAV imagery was most effective for mapping and managing high-density infestations. The cost of drone imagery and data analysis was estimated at $10-$30/ha, depending on the data resolution.

Barley crops showed better competitiveness against barley grass than wheat, with lower weed seed numbers observed in barley paddocks.